5G for Automation in Industry - White Paper Primary use cases, functions and service requirements - March 2019 5G Alliance for Connected ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
White Paper
5G for Automation in Industry
Primary use cases, functions and service requirements
March 2019
5G Alliance for Connected Industries and Automation5G for Automation in Industry Contact: Email: info@5g-acia.org www.5g-acia.org Published by: ZVEI – German Electrical and Electronic Manufacturers’ Association 5G Alliance for Connected Industries and Automation (5G-ACIA), a Working Party of ZVEI Lyoner Strasse 9 60528 Frankfurt am Main, Germany www.zvei.org March 2019 Graphics: ZVEI The work, including all of its parts, is protected by copyright. Any use outside the strict limits of copyright law without the consent of the publisher is prohibited. This applies in particular to reproduction, translation, microfilming and storage and processing in electronic systems. Despite the utmost care, ZVEI accepts no liability for the content.
Contents
1 Introduction 4
2 3GPP 5
3 5G-ACIA 5
4 Areas of application and use cases for automation 5
in manufacturing
4.1 Motion control 8
4.2 Control-to-control 9
4.3 Mobile control panels 9
4.4 Mobile robots 9
4.5 Massive wireless sensor networks 10
4.6 Remote access and maintenance 11
4.7 Augmented reality 11
4.8 Closed-loop process control 12
4.9 Process monitoring 12
4.10 Plant asset management 12
5 The primary functions provided by 5G for factory 13
and process automation
5.1 Quality of service (QoS) for 5G communication services 14
5.2 Data traffic characteristics 14
5.2.1 End-to-end latency 15
5.2.2 Data rate 16
5.2.3 Time synchronicity 17
5.3 Dependability 17
5.3.1 Communication service availability 17
5.3.2 Communication service reliability 18
5.3.3 Dependability and assurance 18
5.4 Deployment 19
5.4.1 Non-public networks 19
5.4.2 Slicing and isolation 20
5.5 Interworking 22
5.5.1 Seamless integration 22
5.5.2 Interworking with 3GPP systems 24
5.6 Security 24
5.7 Positioning 25
5.8 Resource and energy efficiency 27
5.9 Operation and maintenance 27
6 References 28
7 Abbreviations 29
8 5G-ACIA members 30
31 Introduction
This white paper examines how the 3GPP-defined 5G architecture will impact industry, in
particular process and discrete manufacturing. It describes the most relevant use cases, and
the corresponding 5G functions and service requirements. It is consciously an overview,
and does not aspire to be fully comprehensive, for instance spectrum aspects are out of
scope. Spectrum aspects are discussed in 5G-ACIA “5G for Connected Industries and Auto-
mation” white paper [10].
Consideration is primarily given to the needs of automation and robotics. These themes are
naturally closely related to the fourth industrial revolution (also known as Industry 4.0),
and the Industrial Internet of Things (IIoT).
This paper is intended to give ICT professionals a high-level overview of relevant 3GPP-
supported operational technology (OT) use cases. Furthermore, it is designed to give OT
companies (i.e. those companies in industry who will deploy 5G) high-level guidance on
what is supported by the 3GPP-defined architecture, and to provide references where they
can find more detailed information.
The main 3GPP reports related to industrial use cases are as follows:
• Study on Communication for Automation in Vertical Domains (FS_CAV) in Technical
Report (TR) 22.804,
• Feasibility Study on Business Role Models for Network Slicing (FS_BMNS) in TR 22.830
and
• Feasibility Study on LAN Support in 5G (FS_5GLAN) in TR 22.821.
These TRs are not normative (i.e. they are not prescriptive and they are non-binding). They
cannot be used as the basis for implementation, and they cannot be referred to by 3GPP
technical specifications (TS).
It should be noted that 3GPP often employs the term “vertical domain”, i.e. a particular
industry or group of enterprises in which similar products or services are developed, pro-
duced, and provided.
The TRs were primarily written by 3GPP to understand and summarize the high-level com-
munication needs of the industrial community – which have since been described in the
following normative and binding technical specifications (TS):
• Service requirements for cyber-physical control applications in vertical domains,
TS 22.104 and
• Service requirements for the 5G system, TS 22.261.
Once finalized (“frozen”), the TRs given above are not updated. However, TSs are updated
as the need arises. As a result, it is possible for the content of a TS to diverge increasingly
from that of a TR over time. This should be taken into account when making references to
any 3GPP TR/TS.
The functionality described in this white paper is proposed for implementation in 3GPP
Rel-16 specifications, currently scheduled for completion in 2020.
42 3GPP
The 3rd Generation Partnership Project (3GPP) is a collaborative project that brings
together standardization organizations from around the world to create globally accept-
able specifications for mobile networks.
As its name implies, it was first created to establish such specifications for the third genera-
tion (3G) of mobile systems. It has continued its work for subsequent generations, includ-
ing the one considered here, the fifth generation (5G).
3 5G-ACIA
The 5G Alliance for Connected Industries and Automation (5G-ACIA) was established to
serve as the central and global forum for addressing, discussing, and evaluating relevant
technical, regulatory, and business aspects with respect to 5G for the industrial domain. It
reflects the entire ecosystem and all relevant stakeholder groups, ranging from operating
industry (OT) players (industrial automation companies, engineering companies, produc-
tion system manufacturers, end users, etc.), the ICT industry (chip manufacturers, network
infrastructure vendors, mobile network operators, etc.), academia and other groups.
The paramount objective of 5G-ACIA is to ensure the best possible applicability of 5G
technology and 5G networks for connected industries, particularly discrete manufacturing
and the process industry. 5G-ACIA’s mission is to ensure that the interests and needs of
the industrial domain are adequately considered in 5G standardization and regulation.
5G-ACIA will further ensure that ongoing 5G developments are understood by and effec-
tively transferred to the industrial domain.
4 Areas of application and use cases for automation
in manufacturing
Manufacturing is diverse and heterogeneous, and is characterized by a large number
of automation use cases. These can be divided into five distinct areas of application, as
depicted in Figure 1:
1. Factory automation,
2. Process automation,
3. Human-machine interfaces (HMIs) and production IT,
4. Logistics and warehousing, and
5. Monitoring and predictive maintenance.
5Fig. 1: Automation areas in manufacturing
Factory automation
Automation and
manufacturing
Process automation
Human-machine interface (HMI)
Logistics and warehousing
Monitoring and maintenance
Source: ZVEI
Factory automation comprises the automated control, monitoring and optimization of pro-
cesses and workflows within a factory. This includes closed-loop control applications (e.g.
based on programmable logic or motion controllers), robotics, and aspects of computer-
integrated manufacturing.
Example use cases (as described in [1]) for factory automation include motion control,
control-to-control, mobile robots and massive wireless sensor networks. Communication
services for factory automation need to fulfill stringent requirements, especially in terms
of latency, communication service availability and determinism. Operation is limited to a
relatively small service area, and typically no interaction is required with the public net-
work (e.g. for service continuity, roaming, etc.).
Process automation refers to the control of production and handling of substances such as
chemicals, foodstuffs and beverages, etc. The aim of automation is to streamline produc-
tion processes, lower energy consumption and improve safety. Sensors measuring process
parameters, such as pressures or temperatures, operate in a closed loop by means of central
and/or local controllers in conjunction with actuators, e.g. valves, pumps, heaters, etc. A
process-automated manufacturing facility may range in size from a few 100 m² to several
km², or may be geographically dispersed within a specific region.
Example use cases [1] for process automation include mobile robots, massive wireless sen-
sor networks, closed-loop process control, process monitoring and plant asset management.
Communication services for process automation need to meet stringent requirements. For
instance, low latency and determinism are crucial for closed-loop control. Interaction may
be required with the public network (e.g. for service continuity, roaming, etc.).
Human-machine interfaces (HMIs) include many diverse devices for interaction between
people and production systems. These can be panels mounted to a machine or production
line, as well as standard IT devices, such as laptops, tablet PCs, smartphones, etc. In addi-
tion, augmented and virtual reality (AR/VR) systems are expected to play an increasingly
important role in the future.
Production IT, on the other hand, encompasses IT-based applications, such as manufactur-
ing execution systems (MES) and enterprise resource planning (ERP) systems. The primary
6goal of an MES is to monitor and document how raw materials and/or basic components are
converted into finished goods. An ERP system, by contrast, generally provides an integrated
and continuously updated view of business processes. Both systems depend on the timely
availability of large volumes of data from the production process.
Example use cases [1] for HMIs and production IT include mobile control panels and aug-
mented reality systems. Communication services for HMIs and production IT need to meet
stringent requirements. For example, very low latency is imperative for some use cases.
Most HMI and production IT use cases are limited to a local service area, and typically no
interaction is required with the public network (e.g. for service continuity, roaming, etc.).
Logistics and warehousing refers to the organization and control of the flow and storage of
materials and goods in the context of industrial production. Intralogistics is logistics on a
defined premises, for example to ensure the uninterrupted supply of raw materials to the
factory floor by means of automated guided vehicles (AGVs), forklift trucks, etc. Warehous-
ing refers to the storage of materials and goods, for example employing conveyors, cranes,
and automated storage and retrieval systems. For practically all logistics use cases, the
positioning, tracking and monitoring of assets are of high importance.
Example use cases [1] for logistics and warehousing include control-to-control and mobile
robots. Communication services for logistics and warehousing need to meet very stringent
requirements in terms of latency, communication service availability and determinism, and
are limited to a local service area (both indoor and outdoor). Interaction is required with
the public network (e.g., for service continuity, roaming, etc.).
Monitoring and predictive maintenance refers to the monitoring of certain processes and/
or assets, but without immediately impacting the processes themselves (in contrast to a
typical closed-loop control system in factory automation, for example). This includes, in
particular, condition monitoring and predictive maintenance based on sensor data.
Example use cases [1] include massive wireless sensor networks, and remote access and
maintenance. Communication services for monitoring and predictive maintenance are lim-
ited to a local service area (both indoor and outdoor). Interaction is required with the
public network (e.g. for service continuity, roaming, etc.).
The primary manufacturing-domain use cases can therefore be regrouped into the follow-
ing ten categories:
1. Motion control
2. Control-to-control
3. Mobile control panels
4. Mobile robots
5. Massive wireless sensor networks
6. Remote access and maintenance
7. Augmented reality
8. Closed-loop process control
9. Process monitoring
10. Plant asset management
The following table maps the various areas of application to the use case categories.
7Table 1: Areas of application and corresponding use cases
Process monitoring
Remote access and
Augmented reality
Control-to-control
Massive wireless
sensor networks
Motion control
Mobile control
management
maintenance
loop process
Plant asset
Closed-
control
Mobile
panels
robots
Factory automation X X X X
Process automation X X X X X
HMIs and production IT X X
Logistics and warehousing X X X
Monitoring and maintenance X X X X
Source: ZVEI
Each use case category is briefly described below, with examples.
4.1 Motion control
Motion control is one of the most challenging closed-loop control use cases in industry.
A motion control system is responsible for controlling moving and/or rotating parts of
machines in a clearly-defined way, for example for printing presses, machine tools and
packaging equipment. A schematic of a motion control system is given in Figure 2. A
motion controller periodically sends target set points to one or several actuators (e.g. a lin-
ear actuator or a servo drive) that then perform(s) a corresponding action for one or several
processes (in this instance, generally the movement or rotation of a specific component). At
the same time, sensors determine the current state of the process(es) (e.g. the current posi-
tion and/or rotational position of one or multiple components) and send the actual (cur-
rent) values back to the motion controller in a strictly cyclical and deterministic manner.
For example, during each communication cycle, the motion controller sends updated set
points to all actuators, and all sensors send their actual (current) values back to the motion
controller. At present, motion control systems typically employ wired Industrial Ethernet
technologies. Examples of these technologies include PROFINET IRT or EtherCAT – which
support cycle times of less than 50 µs.
In general, motion control has the highest requirements in terms of latency and service
availability. The service areas are usually comparatively limited in size, and no interaction
with public networks is required.
Fig. 2: Schematic of a motion control system
Set
points Act Sense
Motion
controller Actuator Process
Actual values
Sensor
Source: ZVEI
84.2 Control-to-control
Control-to-control (C2C) communication is between industrial controllers (e.g. program-
mable logic controllers or motion controllers). It is already established for a number of use
cases, such as:
• Large items of equipment (e.g. newspaper printing presses), where several control-
lers are employed to cluster machine functions which need to communicate with each
other. These controls typically need to be synchronized, and exchange real-time data.
In general, this use case has very stringent requirements in terms of latency, integrity
and service availability. Typically, to meet the need for low latency and high integrity,
non-public networks are used.
• Multiple individual machines performing a shared task (e.g. machines on an assembly
line) and that often need to communicate with each other, i.e. to control and coordi-
nate the handover of components from one machine to another.
Protocols for today’s C2C communication include Industrial Ethernet standards such as
PROFINET, EtherCAT, OPC UA, and other protocols – which are often based on Fast Ethernet.
C2C communication is expected to increase. In particular, there is likely to be a significant
rise in the number of participating controllers in any given use case and in the volume of
data being exchanged.
4.3 Mobile control panels
Mobile control panels are crucial to interaction between people and production equipment,
and to interaction between people and mobile/portable devices. These panels are mainly
used for configuring, monitoring, debugging, controlling and maintaining machines,
robots, cranes or entire production lines.
In addition, (safety) control panels are typically equipped with an emergency stop button
and an enabling device which an operator can activate when a dangerous situation arises
in order to avoid injury to humans or damage to assets. When the emergency stop button
is activated, the controlled equipment (and possibly neighboring machines) must immedi-
ately be placed in a safe, stationary position.
Similarly, with a “dead man’s switch”, the operator must manually keep the switch in a pre-
defined position. If the operator, for example, inadvertently releases it, the corresponding
equipment must again immediately come to rest in a safe, stationary position.
Due to the critical nature of these functions, safety control panels currently usually have a
wired connection to the equipment. In a 5G radio scenario, a signal must be periodically
sent – and received – in order to verify that the control panel is still connected. The verifica-
tion cycle time always depends on the corresponding process/equipment. For a fast-moving
robot, for example, the cycles are shorter than for a slow-moving linear actuator.
The service area is usually limited in size, as each mobile control panel is associated with a
single, individual item of equipment.
4.4 Mobile robots
Mobile robots and mobile platforms, such as AGVs, are employed widely in diverse use
cases in industrial and intralogistics environments. A mobile robot is essentially a program-
mable machine capable of executing multiple operations, traveling along preprogrammed
routes to perform a large variety of tasks.
9A mobile robot is able, for example, to move goods, materials and other objects, and can
have a significant range of movement within a given industrial environment. Mobile robot
systems are able to perceive their surroundings, i.e. they can sense and react to their envi-
ronment.
AGVs are a sub-group within the mobile robot category. AGVs are driverless vehicles that
are steered automatically. They are employed to efficiently move goods and materials
within a defined area.
Mobile robots are monitored and controlled by a guidance control system. This control
system is required to transmit real-time information to avoid collisions between robots, to
assign tasks, and to manage robot traffic. The mobile robots are track-guided by means of
markings or wires in the floor, or guided by their own surround sensors, such as cameras
or laser scanners.
Mobile robots and AGV systems must frequently interoperate with conveyor assets (cranes,
lifts, conveyors, industrial trucks, etc.), and monitoring and control elements (sensors and
actuators). They also need to exchange data for reporting, e.g. inventories, goods move-
ments and throughput, for tracking and monitoring, and for forecasting. Service areas can
be very large.
4.5 Massive wireless sensor networks
Sensor networks are designed to monitor the state or behavior of a particular environment.
In the context of manufacturing, wireless sensor networks (WSN) monitor processes and
equipment, and the corresponding parameters. This environment is typically monitored
using diverse sensor types, such as microphones, CO2 sensors, pressure sensors, humidity
sensors, and thermometers. These sensors together typically form a distributed monitoring
system.
The data captured in this way is leveraged to monitor diverse parameters, for example to
detect anomalies.
5G has the potential to take these networks to the next level: massive machine-type com-
munication (mMTC) will enable massive wireless sensor networks, featuring millions of
devices per square kilometer, i.e. of a size and density far beyond today’s wireless sensor
networks.
Wireless sensor networks are highly dynamic in nature, changing significantly over time in
terms of the type, number and position of sensors deployed. The position of sensors may
be constrained by the available wireless sensor network hardware. Given that sensors are
typically relatively simple devices, the corresponding functionality usually needs to be
modeled in a centralized computing infrastructure. In certain instances, functionality may
be supported within or shared with the sensor network.
In many scenarios, it is necessary to lay expensive cables to supply electrical power to sen-
sors. Alternatively, the sensors are battery powered. In this latter instance, batteries must
be capable of reliably supporting sensor operation, including the communication module,
for extended periods, even years.
104.6 Remote access and maintenance
Remote access is the ability to establish contact and communicate with a device from a
distant location, and this is often the means for performing remote maintenance. Although
industrial networks are isolated from the internet, remote access is already possible, i.e.
via peer-to-peer communication links between just two devices, fieldbuses with multiple
devices and controllers, LANs or WLANs. However, this requires gateway functionality.
Remote access to device data requires mapping of data formats, addresses, coding, units,
and status at each transition in the automation pyramid (see Figure 3). This can entail
significant engineering effort. Moreover, data mapping implemented in the gateway(s) is
relatively static.
An example use case is the inventorization of devices and periodic extraction of configura-
tion data, event logs, version data, and predictive maintenance information. Generically,
this is known as asset management, and tools for collecting and displaying data from mul-
tiple connected devices are called asset monitors.
A system of this type might operate autonomously (a set of configured periodic checks) or
interact with a user (“show me the status of this device”). The remote diagnostics system
might be operated by, for instance, a manufacturer for devices deployed in its factory. Or
the system might be operated by the device vendor as a service for its customer(s).
Fig. 3: Remote access to devices in existing systems through
controllers via gateways at each transition point of the
automation pyramid
Storage Server
Enterprise
Gateway
Factory
Gateway
Shop floor /
Production cell
Control level
Gateway
Field level Cross layer
data exchange
Product
Source: ZVEI
Traditional Automation Pyramid
4.7 Augmented reality
Augmented reality (AR) is a technology that allows a computer-generated image to be
superimposed on a user’s view of the physical world, providing a composite view.
People will continue to play an important and substantial role in production. But
factory-floor workers, for instance, need effective support, i.e. assistance that allows them
to rapidly become familiar with and adept at new tasks, and that ensures they can work in
an efficient, productive and ergonomic manner.
11In this respect, head-mounted AR devices with see-through displays are especially attrac-
tive since they enable maximum ergonomics, flexibility and mobility, leaving the hands of
workers free. However, if AR devices are worn for prolonged periods (e.g. an entire work
shift), they need to be lightweight and highly energy-efficient. One way to achieve this is
to offload complex processing tasks to the network (e.g. an edge cloud).
AR is expected to play a crucial role in the following use cases:
• Monitoring of processes and production flows
• Delivery of step-by-step instructions for specific tasks, for example for manual assembly
• Ad-hoc support from a remote expert, for example for maintenance or service tasks
4.8 Closed-loop process control
With closed-loop process control, multiple sensors are installed in a production facility, and
each sensor performs continuous measurements. The sensor-captured data are transferred
to a controller which then decides whether and how to operate actuators. Latency and
determinism are crucial.
In closed-loop process control, sensors distributed throughout the production facility con-
tinuously measure typical process parameters such as pressure, temperature, flowrate or
pH value. Harnessing the sensor-captured data, the controllers operate actuators such as
valves, pumps, and heaters/coolers to manage the production process in an optimized, safe
and reliable way.
In these scenarios, determinism and availability are essential as these processes run con-
tinuously over extended periods.
4.9 Process monitoring
With process monitoring, multiple sensors are installed in a production facility to grant
visibility into process or environmental conditions, or into inventories. Data are transmit-
ted to displays for observation and/or to databases for logging and trend monitoring. The
communication service must support high sensor density, and provide low latency and high
service availability. In addition, any battery-driven sensors, including the communication
module, must be highly energy-efficient.
4.10 Plant asset management
To keep a plant up and running, it is essential that assets such as pumps, valves, heaters,
instruments, etc., are well maintained. Timely recognition of any degradation, and ongoing
self-diagnosis, are used to support and plan maintenance work. This calls for sensors that
provide visibility into process or environmental conditions.
Remote software updates modify and enhance components in line with changing condi-
tions and advances in technology.
In plant asset use cases, positioning is an essential requirement. Latency and service avail-
ability are also requirements, but they are less critical than positioning.
125 The primary functions provided by 5G for factory
and process automation
The primary functions provided by 5G for factory and process automation are summarized in Table 2. In the following sections we
will describe those functions in detail.
Table 2: Primary functions provided by 5G for factory and process automation
Functionality Type/ Component Examples
Quality of service Data traffic • Periodic deterministic communication
• Aperiodic deterministic communication
• Non-deterministic communication
• Mixed traffic
End-to-end latency 0.5 ms to 500 ms
Data rate Up to several Gbps
Time synchronicity Down to 1 µs
Dependability Communication service Varies from 99.9 % to 99.999999 %
availability
Communication service
Varies from 1 day to 10 years
reliability
Deployment Non-public networks Standalone: NPN and PLMN are deployed on separate network infrastructure
Hosted: NPN is hosted completely or in part on PLMN infrastructure
Integrated: NPN 5G network is integrated into a larger non-3GPP communication
network such as an IEEE 802 based network
Slicing and isolation MNO provides the virtual/physical infrastructure and virtual network function
• a third party uses the provided functionality or
• a third party manages some virtual network functions via APIs exposed by the MNO
MNO provides the virtual/physical infrastructure
• a third party provides some of the virtual network functions
A third party provides and manages some of the virtual/physical infrastructure and
virtual network functions
Interworking Seamless integration 5G can be integrated with wired technologies on the same machine or production
line
With legacy 3GPP 5G supports mobility between a 5G core network and an evolved packet core (EPC,
the 4G core network)
Security Availability 20 years
Integrity Data received not tampered with and was transmitted by the sender
Confidentiality Optimizing and minimizing signaling overhead, particularly for small packet data
transmissions
In-network caching and operating servers closer to the network edge
Positioning Between 0.2 m and 10 m
Efficiency Spectrum, battery (power)
and protocol efficiency
Operation and • Fault management
maintenance • Configuration management, including provisioning and lifecycle management
• Accounting, including online and offline charging
• Performance management, including the definition of key performance indicators
(KPIs)
• Security management
Source: ZVEI
135.1 Quality of service (QoS) for 5G communication services
For 3GPP, quality of service (QoS, known as communication service performance) com-
prises four parameters: communication service availability, communication service reli-
ability, end-to-end latency, and user-experienced data rate. These are regarded as core 5G
requirements (termed characteristic parameters by 3GPP).
3GPP also defines six secondary requirements (known as influence quantities or param-
eters): message size, transfer interval, survival time, user equipment (UE) speed, number
of UEs, and service area.
3GPP has defined multiple key performance indicators (KPIs) for all the above parameters,
reflecting the specific needs of various use cases.
3GPP also describes further requirements, such as timeliness, positioning and time syn-
chronicity, again with corresponding KPIs.
All the above KPIs are given in TS 22.104 [2], which is specific to factory and process auto-
mation services.
In addition to KPI-defined requirements, 3GPP has described some general QoS service
requirements in TS 22.261 [3] to help 5G support flexible service deployments. This speci-
fication is applicable to all communication services, including those in factory and process
automation.
5.2 Data traffic characteristics
Due to the multitude of use cases, 3GPP has grouped and categorized certain traffic types
based on similar KPIs. 3GPP has so far defined four traffic classes for factory and process
automation. Each traffic class has a specific set of KPIs:
• Periodic deterministic communication: this has stringent requirements in terms of com-
munication service timeliness and availability. For this kind of traffic, 3GPP has defined
KPIs for some key characteristic parameters, such as
• communication service availability,
• communication service reliability measured as mean time between failures,
• maximum end-to-end latency,
• service bit rate/user experienced data rate, and
• time synchronicity.
It has also defined influence parameters associated with these KPIs, such as
• message size,
• periodic transfer interval,
• survival time,
• user equipment (UE) speed,
• number of active UEs, and
• service area.
• Aperiodic deterministic communication: without a pre-set sending time, but still with
stringent requirements in terms of communication service timeliness and availability.
For this kind of traffic, 3GPP has defined KPIs for some key characteristic parameters,
such as
14• communication service availability,
• communication service reliability measured as mean time between failures,
• maximum end-to-end latency, and
• service bit rate/user experienced data rate.
3GPP has also defined influence parameters associated with these KPIs, such as
• message size,
• periodic transfer interval,
• survival time,
• UE speed,
• number of active UEs, and
• service area.
• Non-deterministic communication: this comprises all traffic types other than periodic/
aperiodic deterministic communication. This includes periodic/aperiodic non-real-
time traffic. For this kind of traffic, 3GPP has defined KPIs for some key characteristic
parameters, such as
• communication service reliability measured as mean time between failures, and
• service bit rate/user experienced data rate.
It has also defined influence parameters associated with these KPIs, such as
• UE speed,
• number of active UEs, and
• service area.
• Mixed traffic: this comprises traffic that cannot be exclusively assigned to one of the
other communication types. For mixed traffic, 3GPP has defined KPIs for some key
characteristic parameters, such as
• communication service reliability measured as mean time between failures, and
• service bit rate/user experienced data rate.
And it has defined influence parameters associated with these KPIs, such as
• UE speed,
• number of active UEs, and
• service area.
Taken together, these individual QoS KPIs and influence parameters allow a 5G system to
support a complete solution that meets the QoS needs of factory and process automation.
Certain KPIs, such as end-to-end latency, data rate, and time synchronicity, are presented
below.
5.2.1 End-to-end latency
From the 3GPP 5G perspective, end-to-end latency is the time it takes to successfully trans-
fer a given piece of information from a source to a destination, i.e. between the trans-
mitting 5G communication service interface and the receiving 5G communication service
interface(s), as shown in the following figure. For the purposes of this document, latency is
always one-way latency rather than round-trip latency.
15Fig. 4: End-to-end latency measured between the CSIFs
distributed
5G service performance requirements distributed
automation automation
application application
CSIF CSIF
5G system
Communication system
CSIF – communication service interface between distributed automation application/function and 5G system
Source: ZVEI
In TR 22.804 [1] and TS 22.104 [2], latency generally includes only one wireless link (UE
to network node or network node to UE) rather than two wireless links (UE to UE), except
for certain specific use cases, such as electrical power distribution/automated switching for
power isolation and restoration.
In the latest TR 22.804 [1] and TS 22.104 [2], the lowest maximum latency requirement
for deterministic communication service is defined as 0.5 ms for small industrial environ-
ments (50 m x 10 m x 10 m) with small packet sizes for motion control. A higher maximum
latency of 500 ms is given for periodic communication for standard mobile robot operation
and communication services for CCTV surveillance cameras in mass rail transit with defined
speeds of movement for each UE (UE speed < 50 km/h for mobile robots and ≤ 160 km/h
for mass rail transit in urban environments).
In a 3GPP-defined 5G system, if the packet is not delivered within the defined window
(maximum latency), the packet is considered lost. Within this window, network layer pack-
ets may be retransmitted one or multiple times in order to meet the defined reliability
requirement.
5.2.2 Data rate
TS 22.261 [3] describes 5G requirements from the user perspective, in particular the user
experienced data rate: this is generally given as the minimum data rate required to achieve
a user experience of sufficient quality, with the exception of broadcast-type services, where
the value given is a maximum.
TR 22.804 [1] and TS 22.104 [2] define/specify the service bit rate. The rate is defined
slightly differently for the various traffic classes:
a) Deterministic communication
For deterministic communication, 3GPP defines the user experienced data rate as the
committed data rate requested by the communication service. In TS 22.104 [2], the high-
est data rate requirement for deterministic traffic is given as 10 Mbit/s for mobile robots
with video streaming.
b) Non-deterministic communication
For non-deterministic communication, the rate at which data is transmitted must not fall
below a defined lower limit. This is the required user experienced data rate. The highest
data rate for non-deterministic traffic is given as 1 Gbits for communication between
mechanically coupled train segments.
165.2.3 Time synchronicity
Time synchronization is important for many factory and process use cases as well as for
the 5G network itself. Time synchronicity requirements are given in TR 22.804 [1] and TS
22.104 [2] section 5.6. Fulfillment of these requirements is the basis for processing and
transmitting data in accordance with IEEE 1588v2 (Precision Time Protocol), and for imple-
menting mechanisms to synchronize user-specific time clocks of UEs with a global clock
and/or a working clock. To account for the complexity of the real-world factory floor, 5G
supports up to 32 working clock domains.
TS 22.104 [2] describes three time synchronicity requirements for factory and process auto-
mation. The most stringent requirement is < 1 µs for motion control (up to 300 UEs in a
service area of ≤ 100 m x 100 m) and smart grid power distribution/power management
unit (PMU) synchronization (up to 100 UEs in a service area of < 20 km2). The latter use
case is extremely challenging, and it is probable that 3GPP will have to consider comple-
mentary non-3GPP technologies to achieve this degree of wide area synchronization.
5.3 Dependability
Dependability encompasses availability, reliability, dependability and assurance. These
parameters apply to logical connections (rather than physical). They are explained in detail
in section 4.3 of TR 22.804 [1]. A brief overview is given here.
5.3.1 Communication service availability
Availability is described as the “ability to be in a state to perform as required” [9] or readi-
ness for correct operation. In [9], two detailed definitions are given: network availability
and communication service availability. Communication service availability (CSA) is given
in the requirement tables and the QoS tables in TS 22.104 [2]. CSA is based on the values
for service uptime and downtime. An additional parameter required to calculate CSA is
survival time. 3GPP defines CSA as the “percentage value of the amount of time the end-
to-end communication service is delivered according to an agreed QoS, divided by the
amount of time the system is expected to deliver the end-to-end service according to the
specification in a specific area.”
A detailed description of this relationship can be found in Annex A in TR 22.804 [1] and
Annex C in TS 22.104 [2]. An example for periodic communication is given in Figure 5.
Messages must be sent within a given transfer interval. A figure one in a green field indi-
cates the message has been correctly received. A figure zero in a red field indicates that
the message has been incorrectly received or lost. The example is for a survival time of
three times the transfer interval. This means if two messages in sequence (case 1) are
incorrectly received or lost within the network (NW) and the following message is cor-
rect, the communication service is still considered available. If three messages (case 2) or
more in a sequence are incorrect or lost before receipt of the next correct message, the
communication service is considered unavailable. According to 3GPP’s definition, the CSA
percentage can be calculated over longer periods of time to assess the overall quality of
the connection.
17Fig. 5: Comparison of network (NW) and communication
service (CS) availability depending on consecutively lost
messages
Case 1 Case 2
NW 1 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 0 0 1 0 0 0 0 0 1
CS 1 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 0 0 1 0 0 0 0 0 1
Source: ZVEI
5.3.2 Communication service reliability
Reliability is the “ability to perform as required, without failure, for a given time interval,
under given conditions [9]” or, in other words, continuity of correct operation. Communi-
cation service reliability (CSR) is the mean time between failures (MTBF), calculated using
the cumulative value for time between failures (example values are given in TS 22.104
[2]). It is important to note the differing definitions (Table 3) for reliability in the context
of dependable communication and in the context of TS 22.261 [3], which defines it as a
correctly received message ratio expressed as a percentage. The relationship between com-
munication service availability and reliability in [7] is described in TS 22.104 [2] section 5.
Table 3: Difference between communication service reliability
with MTBF and reliability with correct received packet ratio and
correct received message ratio
Communication service reliability Reliability [7]
Mean time between failures Correct received packet ratio Correct received message ratio
(lower communication layers) (higher communication layers)
~ 1 year 99.3 % 99.999 %
Source: ZVEI
A message consists of several packets. Without taking any corrective measures in case of
packet errors, the correctly received message ratio would be lower than or equal to the
correctly received packet ratio. Due to correction mechanisms applied (e.g. repetition of
packet or forward error correction) the correctly received message ratio can be higher than
the correctly received packet ratio.
5.3.3 Dependability and assurance
When measuring dependability, QoS has a key role to play. Four types of value are classi-
fied in TS 22.261 [3] for each parameter (e.g. CSA, CSR, lost message ratio): required value,
offered value, achieved value, and perceived value. It is important to identify the interface
where each value is measured.
• The required value is defined by the communication service user.
• The network operator (communication service provider) defines the offered value.
• The achieved value is obtained by measurements made by the network operator during
system operation.
• The perceived values are obtained by measurements made by the service user during
system operation.
18The achieved value and the perceived value should be the same, as they are measured at
the same interface. An example for CSA and mean time between failures is given in Table 4.
Table 4: Example of the four types of values for communication
service availability and MTBF
Required Offered Achieved Perceived Unit
value value value value
Communication service
availability
95 95 98.38 98.38 %
Mean time between
failures
30 30 45.25 45.25 days
Source: ZVEI
5.4 Deployment
5.4.1 Non-public networks
In contrast to a public land mobile network (PLMN) that offers mobile network services
to the general public, a 5G non-public network (NPN, also sometimes informally called a
private network) provides 5G mobile network services to a clearly defined user organization
or group of organizations. This organization – also called a vertical domain – deploys the
5G non-public network on its premises, such as the factory floor, an industrial plant, or for
its installed devices (e.g. within a smart grid).
The user organization administers the non-public network itself; the network can be oper-
ated by the user organization itself, or by a contracted service provider.
For vertical domains, non-public networks can be desirable for several reasons:
• High QoS requirements: Communication services for automation in industrial networks
are associated with high QoS requirements, especially in terms of availability and
latency. These cannot be satisfied by the public mobile network. It is very probable that
other public-network communication services, such as voice calls, video and internet
traffic, with their much lower QoS requirements, would negatively impact the vertical
communication services used for automation.
• Dedicated security credentials: Industrial communication networks need security
credentials that are different from those used in public networks. The same credentials
are often used for both device authorization and communication encryption. For some
industrial use cases, the credentials are subject to regulation.
• Isolation from other networks: The industrial communication network needs to be
isolated from the public mobile network for reasons of performance, security, privacy,
and safety.
• Accountability: A non-public network makes it easier to identify responsibility for avail-
ability, maintenance, and operation.
A number of industrial use cases (mentioned in section 4) require service continuity across
and between the NPN and the PLMN. This might, in certain instances, necessitate roaming
between the NPN and PLMN. It is important to note that service continuity when interoper-
ating with public networks requires security credentials for and the agreement of the public
network for the corresponding user devices.
19Three deployment scenario categories for non-public networks have been defined by 3GPP:
standalone, hosted, and integrated [16]. Table 5 shows the three categories according to
[16].
Table 5: Deployment scenario categories for non-public networks
according to [16]
Standalone:
NPN and PLMN are deployed on
separate network infrastructure.
NPN
Distributed Distributed
automation
application NPN automation
application
Hosted:
NPN is hosted completely or in
part on PLMN infrastructure,
PLMN
implemented by means of e.g.
software, policies and/or network
slicing (private slice). Distributed Distributed
automation
application NPN automation
application
Integrated:
NPN 5G network is integrated into
a larger non-3GPP communication Communication network
network. The NPN 5G network is
a subset of the communication
infrastructure. The NPN 5G network Distributed Distributed
provides, for instance, wireless automation
application
Non-3GPP
network NPN Non-3GPP
network
automation
application
links.
Source: ZVEI
Many of the vertical use cases given in TR 22.804 [1] assume the deployment of non-public
networks for the above-mentioned reasons. This is also true for the manufacturing use
cases described in section 4. Normative 5G service requirements for non-public networks
are given in section 6.25 of TS 22.261 [3].
5.4.2 Slicing and isolation
Network slicing is one of the key features of 5G. It allows the network operator to pro-
vide customized networks for specific services, and to achieve varying degrees of isolation
between the various service traffic types and the network functions associated with those
services.
For example, within a single network infrastructure, there may be diverse needs in terms of
functionality (e.g. priority, charging, policy control, security, and mobility), in performance
(e.g. latency, mobility, availability, reliability and data rates), or operational requirements
(e.g. monitoring, root cause analysis, etc.).
Moreover, the slices can be configured to serve specific user organizations (e.g. public
safety agencies, corporate customers, roamers).
20A network slice can deliver the functionality of a full-fledged network, including radio
access network and core network functions (potentially from multiple network equipment
vendors). One network can support one or several network slices.
Network slicing also opens up new ways of isolating and separating data traffic types, e.g.
in a factory. Moreover, slicing can be employed to isolate the various logical networks, as
mandated by IEC 62443 [17].
According to 3GPP, each item of 5G user equipment is able to support up to eight slices
for multiple logical networks carrying diverse data types e.g. transferring monitoring data
to an MES via OPC UA, and at the same time supporting Industrial Ethernet traffic for the
control of devices by a PLC.
3GPP TS 22.261 [3] section 6.1 describes some basic service requirements that 5G must
fulfill to enable network slicing:
• Network slicing management by the operator, i.e. creating, scaling, modifying,
configuring, deleting slices;
• UE management for the network slice,
• Traffic isolation and network resource management within and between network slices,
• Roaming, and
• Cross-network slice coordination.
5G is an enabler of new business models. There is therefore strong interest from verticals,
such as the factory and process automation industry, in leveraging network slicing to allow
new business relationships between mobile network operators (MNOs) and OTs. 3GPP TR
22.830 [5] addresses these relationships and the corresponding requirements.
TR 22.830 [5] considers a number of potential new business models, and these will be used
as the basis for further work. There are four potential business models:
• Model a: the MNO provides the virtual/physical infrastructure and virtual network func-
tions; a third party uses the functionality provided by the MNO,
• Model b: the MNO provides the virtual/physical infrastructure and virtual network func-
tions; a third party manages some virtual network functions via APIs exposed by the
MNO,
• Model c: the MNO provides the virtual/physical infrastructure; a third party provides
some of the virtual network functions,
• Model d: a third party provides and manages some of the virtual/physical infrastructure
and virtual network functions.
To facilitate these models, 5G supports the following [3]:
• Hosting of private slices by an MNO for a defined user organization, e.g. an OT
• Network openness (APIs) to expose network capabilities to and enable management by
the user organization within its slice(s)
• Enhanced security, including authentication and authorization, for non-public networks
and private slices hosted by MNOs.
215.5 Interworking
5.5.1 Seamless integration
For most communication on the factory floor and enterprise level (see Figure 6), 5G can be
employed as soon as the 5G spectrum and products are available. In current factories, all
time-critical use cases (e.g. controllers and field devices) have wired connections. Wireless
is mainly used for non-critical services. 5G networks for wireless automation on the factory
floor support a wide variety of sensors, devices, machines, robots, actuators, and terminals.
These may be directly connected to public or non-public networks and/or be connected via
gateway(s).
Fig. 6: Predominant use cases of TCP/IP and Ethernet traffic
Latency Reliability
Storage Server -50 ms 99.9%
Enterprise
Gateway
Factory
Gateway
Shop floor /
Production cell
Control level
Gateway
Field level Cross layer
data exchange
Product
Traditional Automation Pyramid 99.9999%
Source: ZVEI
A number of Industrial Ethernet technologies (see IEC 61158-1 [13], IEC 61784-2 [14])
are used in industrial automation for real-time communication between controllers and
machines. Compared to standard Ethernet communication, these technologies have very
specialized requirements:
• Layer-2 switching (MAC, Ethertype, VLAN)
• Short cycle times, and support for multiple parallel cycles
• Short data frames (e.g. encapsulated into Ethernet frames)
• High synchronization requirements
• Transfer of functional safety protocols (see IEC 61784-3 [15])
• Machine cloning in order to use the same machine type in a cell / on a line.
Today, in addition to Ethernet-based industrial standards, there are also fieldbuses using
various means of communication, for example Profibus (DB/PA) [13] and Foundation field-
bus H1 in the process industries. Gateways and converters will be used to interface digital
signals to support successful incorporation of 5G technology into industrial networks. 5G
allows the IT and OT networks to be integrated.
22Fig. 7: Example of an Industrial Ethernet network including 5G links for motion control
Distributed automation function Distributed automation function
Application layer interface
communication system
Appilcation layer Appilcation layer
Transport Transport
performance
layer layer
Example: CSIF for CSIF CSIF 5G L3 network (IP)
smart manufacturing
IP layer IP layer
Example: CSIF for CISF CISF
real time process control MAC sub-layer MAC sub-layer 5G L2 network
and motion control
industrial Ethernet (Ethernet)
Physical layer Physical layer
CSIF: Communication
service interface
Mobile radio channel
Source: ZVEI
The IEEE Time Sensitive Networking (TSN) Task Group is developing a TSN standard with the
goal of enabling the delivery of deterministic services via IEEE 802 networks. TSN provides
many of the services needed in factory automation use cases, e.g., time synchronization,
as well as ultra-reliability through redundancy. The aim of TSN is to support both real-time
and non-real-time traffic multiplexed on Ethernet. 5G will therefore support TSN function-
ality. Furthermore, 5G acts as a TSN bridge, supporting seamless interworking with Indus-
trial Ethernet and TSN-based industrial networks.
It is important to address mobility when integrating TSN and 5G wireless networks [2].
Mobility is key to flexibility in the manufacturing process, i.e. making it possible to add
certain manufacturing resources on-demand, for example by relocating a machine to the
corresponding production line. This means that machines that are synchronized with diver-
gent working clock domains may need to interact.
5G supports [1] seamless integration into the existing (primarily wired) infrastructure. For
example, 5G can be combined with wired technologies on the same machine or production
line.
To support seamless migration, the 5G system will support highly deterministic cyclic data
communication and guarantee bounded delays [11]. In addition, it supports seamless
handover between two base stations without any observable impact on the use case, in
particular with regard to safety.
There are many use cases that depend on Layer 2 Ethernet-based communication [4]. Sen-
sors and actuators use non-IP transport services (e.g. Ethernet) to transmit control signals
in legacy LANs at industrial factories. For this scenario, the 5G-LAN service supports com-
munications between UEs via Ethernet-based protocols (i.e. non-IP packets), while ensuring
the defined QoS (e.g. reliability, latency, and bandwidth).
235.5.2 Interworking with 3GPP systems
As mentioned above, 5G systems can be operated as non-public networks to allow deploy-
ment of networks within a factory or plant isolated from public mobile operator networks.
Flexible interfaces allow seamless interoperability and seamless handovers between non-
public and 5G public mobile operator networks.
Furthermore, 5G supports mobility between a 5G core network and an evolved packet core
(EPC, the 4G core network) with minimum impact on the user experience (e.g. QoS).
However, there are still some limitations regarding seamless handover. Seamless handover
between 5G and 2G and between 5G and 3G networks will not be supported [3].
5.6 Security
Most security concepts and solutions were developed for office IT systems and applications.
Security for automation systems in vertical domains is associated with different priorities,
and with different management and operational characteristics and requirements. The fol-
lowing presents a bird’s-eye view of the salient differences between office IT systems and
automation systems.
The three key attributes of security are confidentiality, integrity, and availability (with
availability meaning access to the system in question). While office IT typically prioritizes
confidentiality over integrity, and integrity over availability, the complete opposite is true
for automation in many vertical domains (see Figure 8).
Fig. 8: Differences in security priorities between office IT
and automation
Office IT priority
Availability Integrity Confidentiality
Automation priority
Source: ZVEI
In other words, availability is the main concern for automation security, followed by integ-
rity; confidentiality generally has the lowest priority. The real-time behaviour of automa-
tion systems can also be critical (especially for control use cases).
Automation systems not only put the emphasis on availability – this availability has to be
guaranteed for component lifetimes (as much as 20 years or more) that are typically five to
seven times those of components within office IT environments.
Another significant difference is the security “culture”. Security patches are released and
distributed in relatively long cycles in automation, in particular due to regulatory con-
straints. Additionally, in many industrial environments patches can only be installed during
scheduled maintenance windows. Anti-virus programs are relatively rare in automation,
24as the virus signatures would have to be updated regularly, causing downtime. Instead,
whitelisting can be employed to ensure only authorized, unmodified applications can be
executed. While security testing and auditing is the norm for critical IT infrastructure such
as service centers, security practices are still evolving for automation in vertical domains.
A reason for this lag is that the reliable operation of the automation system must not be
endangered by penetration tests.
Security standards are already well established for office IT systems, but are still under
development for automation systems. However, the industrial security standard IEC 62443
[17][18] is increasingly being adopted for automation systems.
While confidentiality is generally of high importance in office environments, it is typically a
low to medium priority for physically access-restricted automation systems, such as produc-
tion cells in a factory. However, the confidentiality of business-relevant knowledge needs
to be protected, e.g. engineering data or the parameters of a chemical production process.
Integrity is essential and, as already mentioned above, so are availability and reliability.
Non-repudiation may also be a high priority where the correct operation of production
systems is vital, including audit-proof logs, e.g. in the pharmaceutical industry, while it is
typically of medium importance for office IT.
In order to ensure the necessary security, 3GPP has taken into account relevant non-3GPP
specifications, e.g. from ISO and IEC. The main security requirements are given in sections
6 and 8 in [1]. These describe the means of assuring data received have not been tam-
pered with and were transmitted by the sender, i.e. integrity has been maintained, and the
sender cannot deny having transmitted the data (non-repudiation). In addition, section 6
in [4-5] describes special deployment scenarios, such as private networks or network slices
either hosted by a traditional mobile operator, or hosted by an OT company on their own
infrastructure.
5.7 Positioning
Highly accurate positioning is essential for factory and process automation. It is becom-
ing increasingly important to track mobile devices and mobile assets in order to improve
processes and increase flexibility in industrial environments. However, positioning require-
ments in these environments vary considerably.
In some instances, accuracy to within a few centimeters is needed, in others to within sev-
eral meters. In some cases, maximum permissible latency can be measured in milliseconds.
In others, it is sufficient to receive a few position updates each day. In general, automation
and control systems typically need sub-meter accuracy. For tracking, routing and guiding,
accuracy above a meter is sufficient. For certain use cases, such as AGVs and HGVs, precise
positioning is crucial.
The positioning requirements from about 60 use cases (mobile control panels, autonomous
driving systems, modular assembly areas, augmented reality and storage of goods) were
considered in 3GPP ([1][3][6]) and were grouped into seven service levels, suitable for
all positioning use cases between 20 cm and 10 m accuracy. Some of these use cases are
shown in Table 6. The grouping can be found in [3], and includes horizontal and vertical
accuracy, availability, latency, mobility and coverage.
25You can also read
